HACIT: A Privacy Preserving and Low Cost Solution for Dynamic
Navigation and Forensics in VANET
Kevin Decoster and David Billard
University of Applied Sciences Western Switzerland in Geneva - HES-SO, Geneva, Switzerland
Keywords:
VANET, D2D, Distributed Systems, Ad-Hoc network, Privacy, Blockchain, Forensics.
Abstract:
The current architecture for VANET related services relies on a Client-Server approach and leads to numerous
drawbacks, such as network congestion due to the bottleneck problem or, more importantly, data privacy
concerns. Indeed, because of the network topology, traffic must go through nodes which limit the bandwidth
and thus bound the overall network capacity. Finally, user data is collected and stored in servers, used by third
party services. However, these providers are known to treat lightly user privacy by selling or using the data for
their own purposes (Beresford and Stajano, 2004). By use of a decentralized and distributed communication
protocol (i.e. D2D), one can overcome these problems by spreading the communication burden to all nodes in
the mesh. By means of cryptographic techniques, we can ensure that the shared data is secured and controlled
at the end-user side. This paper presents a study and proposes a proof of concept of a decentralized and
distributed information system by means of a dynamic navigation system for VANET, using a low-cost solution
such as Wifi or LTE-direct new 3GPPP protocol. This system preserves user privacy and is augmented with
forensics capabilities.
1 INTRODUCTION
While many vendors offer IoT devices or propose cars
with embedded networked devices, the notions of data
privacy and security are often considered as a minor
point, when the list of the device features is enume-
rated. However, law enforcement services, specia-
lized services, and digital forensic experts witness a
dramatic surge in criminal data leaks, espionage, so-
cial engineering exploits and abuse of industrial con-
trol systems (SCADA). The last semi-annual report
from MELANI (Melani, 2006), the Swiss agency de-
aling with cybersecurity within the Federal Intelli-
gence Service (NDB / SRC), stresses the importance
of handling the security vulnerabilities. At the EU
level, member states have consistently tried to imple-
ment legal frameworks to protect the processing of
personal data and the free movement of such data
within EU (Directive, 1995) and in April two year
ago, a directive imposed a binding legal framework
(EU, 2016). Therefore, all this constitutes sufficient
warning to alert the security community and to trig-
ger advanced research. The research presented in this
paper concerns the protection of user privacy but also
the necessity for law enforcement to gather evidence
(”digital forensics”).
The proof of concept presented in this paper fo-
cuses on the collaboration of intelligent cars for de-
termining the best driving route and avoiding traffic
jams using only local information transmitted by the
other cars, without using a central Internet service. By
forbidding the use of centralized services like Google
Maps or Tomtom Go Mobile, the user private data is
kept at the user’s premises and privacy is strengthe-
ned. For that purpose, we designed a decentralized
and distributed communication scheme for Vehicular
Ad-Hoc Network (VANET) with forensics capabili-
ties. In order to (1) maintain data privacy at the user
side and (2) avoid to rely on costly and embedded har-
dware such as embedded radar or computers in cars,
we focus on a low-cost solution using mobile to mo-
bile device communications (D2D) on Android de-
vice, and validate our model on numerical simulation
aided with a VANET simulator such as GAMA (Grig-
nard et al., 2013) or Veins (OMNET++ and SUMO)
(Veins, 2011).
The modeled system can be decomposed into
three layers:
1. The communication layer, which ensures Confi-
dentiality, Integrity, and Availability (CIA) using
symmetric and asymmetric cryptography (end to
end). Furthermore, we must address the problem
454
Decoster, K. and Billard, D.
HACIT: A Privacy Preserving and Low Cost Solution for Dynamic Navigation and Forensics in VANET.
DOI: 10.5220/0006778404540461
In Proceedings of the 4th International Conference on Vehicle Technology and Intelligent Transport Systems (VEHITS 2018), pages 454-461
ISBN: 978-989-758-293-6
Copyright
c
2019 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
of access control (Identification, Authentication,
and Authorization) in a strong dynamic environ-
ment and anonymity (e.g. by the use of temporal
Id’s);
2. The message dissemination layer: once the secure
communication channel is established, the vehi-
cles (nodes) must send information about the traf-
fic, current attributes, etc., while addressing the
problem of network overload;
3. The application layer, for the processing of the
gathered data. The two applications foreseen in
our testbed are (1) the rerouting using an off-line
partially stored map (i.e. OpenstreetMap API)
and (2) the forensics capability using a blockchain
method.
Furthermore, confidentiality between nodes is en-
sured by asymmetric and symmetric cryptography
while authentication and forgery-proof logging will
be ensured by blockchain technology.
The technology enabling communication is a fra-
mework called GRCBox (Tornell et al., 2015), in-
stalled on a Raspberry Pi device, which allows ad-
hoc networking on Android phones. This paper lea-
ves aside the study of the new proposed technology
ProSe (Prasad et al., 2014), an LTE based techno-
logy which directly compete with the Dedicated Short
Range Communication (DSRC) algorithm used for
VANET networking. It was proposed in Release 12 of
the 3GPP specification for device to device commu-
nication using licensed spectrum. While the deploy-
ment is still in progress, we foresee that this techno-
logy can enhance our proposed algorithm. More ge-
nerally, the communication medium capability is a
key research issue in our work.
The rest of the paper is organized as follow.
Section 2 expresses the current state of the art regar-
ding the proposed functionalities. Section 3 presents
the chosen system models and answers the problem
of communication, message dissemination and dyna-
mic rerouting. Section 4 describes the forensics im-
plementation and section 5 concludes the paper.
2 RELATED WORK
The first layer of our system is the creation of an ad-
hoc network between in-range nodes (vehicles). The
targeted application (i.e. VANET communication) le-
ads to highly dynamic and changing network, mea-
ning a node is expected to connect and disconnect to
a particular network fast. Two cars moving in op-
posite directions at 50km/h, with a WIFI range of
100m leads to a gross calculus of a maximum trans-
mission duration of around 7s. This means that, in
order to connect two nodes, one cannot take longer
than a dozen seconds to perform all connexion stages.
As a consequence, we cannot use the Wifi-Direct op-
tion, available on stock Android smartphone device,
as explained in (Funai et al., 2015). Although too
slow and power consumer, a solution for rooted de-
vice, the Serval project (Serval project, 2016), provi-
des a full mesh network without external hardware.
From the authors’ knowledge, only the GRCBox fra-
mework enables a free and open-source solution for
ad-hoc network deployment for Android devices by
using an external embedded hardware such as a Rasp-
berry Pi. However,they are recent crowd-sourced so-
lution such as GoTenna (GoTenna, 2016) that propose
a proprietary dedicated hardware for mesh networ-
king based radio, which is also under consideration.
Regarding decentralized and distributed device to
device communication, we use the solution proposed
in (Kaul et al., 2017), focusing on efficient and dyna-
mic message dissemination in ITS, to draw our com-
munication model, and avoid the broadcast storm pro-
blem while efficiently delivering traffic information
messages. More specifically to our rerouting appli-
cation, the paper (Leontiadis et al., 2011) proposes
a dynamic routing application. However, they don’t
address the limitations of communication within VA-
NET and they assume the use of external infrastruc-
tures. As a matter of fact, the more recent paper
(Garip et al., 2015) proposes a suboptimal offline re-
routing solution while addressing the communication
problems that might arise in VANET (i.e. request and
response traffic information approach). Nonetheless,
the security and forensic are not fully addressed.
To the best of our knowledge, although the secu-
rity in VANET is a well researched field ((Raya and
Hubaux, 2007)), no paper fully addresses the foren-
sics concern. Often, they assume a third party TA, for
real identity recovery in a secure and anonymous net-
work. However, trusting an external TA would break
the proposed project requirements. This is where
forensics logging comes into play: the whereabouts
and exchanged communication are logged into the
own user’s system while ensuring unforgeability, via
blockchain. For instance, (Sharma et al., 2017) and
(Leiding et al., 2016) use blockchain in VANET. Ho-
wever, they use it for monetary applications such as
an automatic smart contract for insurance or tolling.
The most used blockchain network enabling smart
contract alongside cryptocurrency is Ethereum (see
(Wood, 2014)). However, without the need for a cyp-
tocurrency support (and thus proof of work through
mining), our blockchain can achieve consensus via
randomness and thus avoiding the computational bur-
HACIT: A Privacy Preserving and Low Cost Solution for Dynamic Navigation and Forensics in VANET
455
den of mining. Therefore, our application will be
designed on top of a permissioned blockchain fra-
mework called Hyperledger fabric (linux Foundation,
2016).
3 SYSTEM MODELS
In this section, we present the model design of our
decentralized system.
3.1 High Dynamic Ad-hoc Network
The GRCBox framework installed on a Raspberry Pi
3 model B, with 2 Wifi interfaces, enables ad-hoc net-
working. This architecture allows to communicate
with other cars or with Internet via outer wireless in-
terfaces and to the application via the inner interface
remotely (Wifi). The framework handles the messa-
ges received and forwards them to every outer Wifi
interfaces of nodes connected to the ad-hoc network.
The authors of (Hadiwardoyo et al., 2017) uses the
GRCBox to enable communication between smartp-
hones in order to test a simple application and asses
the performance of the framework. Interestingly, this
system can deliver message at 10HZ, up to 80 meters
and within a delay from 100ms to 900ms. Thus, it
drives us to conclude that this system is really suita-
ble for our practical implementation.
Finally, on the application layer, we propose an
end-to-end secure communication channel and au-
thentication, using the asymmetric cryptography and
the blockchain (see Section 4).
3.2 Scalable Message Dissemination
Each vehicle stores a weighted graph G, containing
for each road segment a weight representing the time
needed to travel this segment and used for offline
shortest path computation (we are using the well-
known Dijkstra (Dijkstra, 1959) algorithm). Alongs-
ide this graph, it stores also a database DB (as an
Hashmap) containing information measured as the
vehicle goes through roads or resulting from the union
of other nodes databases. Each entry is of the form
< RoadId,AvgSpeed,Timestamp >. RoadId is a uni-
que key, meaning that if a new entry is available for
this RoadId, the system will simply keep the latest
update. AvgSpeed is the measured speed on the road
segment with id RoadId. Finally, Timestamp stores
the time at which the speed sample was measured.
Whenever a vehicle joins a network, it sends its
own DB to every node in the joined network. As a
(a) Step one
(b) Step two
(c) Step three
Figure 1: An example of data dissemination.
reply, the receiving nodes send all entries in its da-
tabase whose road ID differs from DB or are simply
more recent.
With this approach, we obtain a simple but effi-
cient 1-hop broadcasting scheme (or also called 1-
persistent broadcast), which avoids the ”broadcast
storm” problem in highly dense traffic. An example of
the dissemination technique for our delay resistant ap-
plication can be found in figure 1. In subfigure 1a, we
observe a car accident at crossroad A, while a green
car comes in the opposite direction and proceeds to
the database exchange with the stopped cars. In sub-
figure 1b, the green car exchanges its database (with
entry at road AB: AvgSpeed = 0) with the blue car.
Finally, in subfigure 1c, the blue car updates its route
and reroutes its course, and thus, it avoids the traffic
jam in crossroad A.
3.3 Dynamic Navigation Rerouting
The rerouting algorithm proposed in this paper uses a
weighting schemes where weights are stored locally
(as a graph) and later used with an optimal shortest
path algorithm.
Weighting Scheme. We recall that the weights of
the roadmap graph are simply computed as the
time needed to travel the road segment. In other
words, given the road segment between edges i
and j, the weight w
i, j
is computed as w
i, j
=
l
i, j
v
i, j
,
with l and v being the road length and average
speed respectively (information available on the
road graph, alongside other parameters such as the
number of lines, type of road etc.). Given a new
estimated average speed ˆv
i, j
measured at time t,
and thus, the corresponding weight ˆw
i, j
, we can
compute the new weight as in Equation 1, with
VEHITS 2018 - 4th International Conference on Vehicle Technology and Intelligent Transport Systems
456
T
2
being a time threshold after which the data is
considered unreliable.
w
i, j
n
= α ·w
i, j
n1
+(1α)· ˆw
i, j
with α =
t
T
2
. (1)
Procedure. Proceed to a simple navigation rerou-
ting, based on newly received traffic information
from other nodes, which can be found in Algo-
rithm 1 and summarized as follow:
After receiving a message containing the gathe-
red road informations db
r
from another node data-
base, where each entry contains a structure of the
form < RoadID,avgSpeed,Timestamp >, we can
distinguish three steps: The road segment weight
update, the navigation path rerouting computation
and the database update.
1. For each entry in db
r
, we update the entry of the
weighted graph G corresponding to the same
road segment (i.e. with same RoadId) using
Equation 1. Then, we verify (1) if the diffe-
rence of speed is bigger than a certain threshold
(i.e. T
1
) (2) if it has not expired regarding a time
threshold T
2
and (3) if the road segment path list
R contains this road segment and we store the
boolean result in a global variable.
2. If the variable containing the result of the pre-
vious step tests is true, we must recompute the
shortest path (using Djisktra) given the new up-
dated graph G, the current position and the des-
tination (i.e. R.end).
3. Finally, we merge the received set db
r
to the
database DB containing the traffic information
sent by other nodes and recorded by the user,
keeping only the latest database entry.
The full algorithm is shown in Figure 1.
4 PRIVACY AND FORENSICS
Section 3 presents a simple but complete communi-
cation model in order to connect nodes in communi-
cation range and exchange packet frames. Once the
traffic information is filtered and stored in each node,
it is processed to update road weights in the roadmap
graph used in another run of Dijkstra algorithm to find
the new shortest path, if different. However, the in-
novative approach in this work is to tackle the pro-
blem of a complete decentralized and secure system
while enabling forensics capabilities. We can divide
the problem into (1) the forensics system and (2) the
security between nodes.
Algorithm 1: Navigation rerouting algorithm upon newly
received traffic information.
Require: db
r
,R, G.
1: db
r
is a list containing a structure of
[RoadID,AvgSpeed,Timestamp], R and G
are the road path list and the road weighted graph
respectively
2: function PATH REROUTING(db
r
,R, G)
3: isModi f y False
4: for d in db
r
do
5: t d.Timestamp
6: w G[d.RoadID].Weight
7: v d.AvgSpeed New speed
8: n current time
9: G[d.RoadID] Updated as Eq 1
10: isModi f y ||w v|| T
1
& ||t n|| T
2
& R.contains(d.RoadId)
11: if isModify then
12: R DJISKT RA(G, position,R.end)
13: DB DB db
r
Keep latest information
4.1 Forgery-proof Logging with
Blockchain
Our original contribution to the VANET community
is to propose a decentralized VANET communication
network without relying on any external third party.
However, the need to provide evidence might arise,
such as in car accident or police investigation (eg: a
previous car behavior), in order to establish respon-
sibilities. Due to the problem specification, such as
low bandwidth capacity and computation on a mo-
bile device, we propose that the user stores in its lo-
cal storage blocks containing logged data and chained
with hash, similar to a blockchain. The hashing of
data guarantee a one way encryption function, mea-
ning it guarantees that the data is untouched as long
as the hash submitted is immutable. Obviously, we
don’t want the user to modify this data considering
forensics fraud. Therefore, we propose the following
scheme (see Figure 2):
A user initially submits a smart contract (also
known as chaincode) similar to Figure 3, which
contains a public hash, the owner public key and
a public function executable only by the owner
of the Smart Contract, to the network consensus.
This step is used as registration. In other words,
once the smart contract code is compiled, the new
user needs to submit the contract to the other peers
for consensus. Upon validation, the user recei-
ved a confirmation alongside the hash of his smart
HACIT: A Privacy Preserving and Low Cost Solution for Dynamic Navigation and Forensics in VANET
457
Figure 2: Blockchain pipeline.
contract (i.e. contract index). An example of an
initiating smart contract is shown in 3.
While the user uses the system, it logs a block of
data (i.e. an offline loop) and computes the hash
of the newest block (i.e.: blockchain).
When user connect to network, it sends the latest
hash alongside its database (road informations).
The receiving node returns its database as well as
the signed hash.
When a user is connected to the Internet through
Wi-Fi or cellular network, it executes the setHash
function given its accumulated hash (signed or
not).
Every time a node is connected to the network, it
downloads the new blockchain version (i.e. up-to-
co n tr a ct Cli e nt_ c on t r ac t {
by te s [] has he s = ne w b yte s ( ) [ N ];
ui nt [] v a li d at i on = n ew uint () [N ];
ad d re ss pub li c pk = m sg . s en de r ;
fct s e th a sh ( b yt es [] sH ash , ad d re ss [ ] oPk ) {
re q ui re ( m sg . s en de r == this . pk )
;
re q ui re ( sH as h . le ngt h == o Pk . l en gt h );
for ( u in t i = 0; i < sH as h . len gt h ; i ++)
{
# no v a li d at i on for t hi s ha sh
if ( oPK [i] = = n ul l )
{
th is . h as hes . push ( h );
th is . v a li d at i on . p us h
(0) ;
}
el se {
by te [] h = S H A2 56 ( s Has h [ i ]) ;
by te [] h_ p = de c ry pt ( de cry pt (
sH as h [ i ] , o Pk [ i ] ) , pk ) ;
if ( h == h_p ) pu sh ( s Ha sh [ i ])
}
}
}
fct p us h ( by te s a )
{
if ( ha she s . co n ta i ns ( a )){
ui nt i = has he s . i nd exo f ( a ) ;
ha she s [ i ] = a ;
val ida t io n s [ i ] += 1;
}
el se {
ha she s . pu sh (a ) ;
val ida t io n s . pu sh ( 1) ;
}
}
}
Figure 3: Proposed initial Ethereum smart contract.
date hash from all nodes).
This scheme allows the user to verify, using the
hash available in his or her smart contract, that the
data he/she owns is valid and not modified. Moreover,
encountered nodes sign the latest hash which add a
layer of proof. A legal entity can then check the ledger
to see which submitted hash were endorsed by other
peers.
Without the need for cryptocurrency (i.e. mone-
tary scheme), the proof-of-work is useless here. Thus,
the Byzantine fault tolerance consensus algorithm can
be achieved by simply selecting a peer ”leader” rand-
omly, which will gather all transactions (smart con-
tract execution) into a new block. As for most block-
chain technologies, this block is then broadcast even-
tually to all peers with the gossip protocol. Therefore,
such a consensus algorithm is considered computati-
onally easy and suitable for mobile application.
VEHITS 2018 - 4th International Conference on Vehicle Technology and Intelligent Transport Systems
458
4.2 Inter-node Security
Furthermore, thanks to the scheme, a receiver node
can retrieve the public key of the sender, given the
sender’s smart contract index (i.e. smart contract hash
address, considers unique among users and serving to
identify), in the blockchain. An example of the ex-
change can be seen in figure 4. In other words, we
guarantee that the data remains confidential within
our service users (i.e. valid registered users) while
being able to authenticate data and avoid malicious in-
jection, by means of asymmetric encryption. Indeed,
a node A maintains an up-to-date and encrypted ver-
sion of its database C, containing gathered road traffic
informations, using his private key K
a
. This encryp-
ted data is computed at a certain interval considering
a trade-off between power consumption and the no-
velty of the data. Once it has joined a new network,
it broadcasts its encrypted data C to all other nodes
alongside its blockchain identity H
s
c (smart contract
address). Upon message reception, the receiving node
uses the unencrypted As id H
s
c to retrieve As public
key PK
a
into the permissioned ledger that it has alre-
ady access to. If no address matches or no public key
available, A is not registered to the network. Other-
wise, we use PK
a
to decrypt C. A challenge is finally
tested to assess the validity of the data.
By means of blockchain functionalities, we pro-
pose a simple way of encryption and authentication.
Of course, it implies that the user needs to have an up-
to-date ledger in order to acknowledge newly subscri-
bed users and retrieve their public key, and that the
data is secured and validated by the community.
4.3 Malicious Node
The proposed model allows for an user to prove the
data provenance and its integrity based on validation
from other encountered users. However, a malicious
node could alter its data while still submitting a valid
hash. Thus, we propose the following extensions:
White-box cryptography (Brecht, 2012): While
this field is still under study, it proposes a way
to hide private key into the source code. In our
case, it could secure the submitted hash, providing
a way for any nodes to verify that it has been sig-
ned by our application (and not submitted a pos-
teriori).
50% overlaps: Each block of data will overlaps
each other, to help check data consistency at the
forensics level.
However, the incentive to submit as much valid
and correct hashes as possible in order to protect an
Figure 4: Authentication and privacy pipeline.
user legally will eventually protect against malicious
behaviour.
5 CONCLUSION
Using well known techniques, we propose a real-time
low-cost system able to connect vehicles in a network
and exchange traffic information in order to dynami-
cally reroute user to spare overall time. With a sim-
ple raspberry pi and GRCbox framework, we enable
multi interface short-range communications through
an ad-Hoc network. We opt for a simple 1-hop bro-
adcasting technique in order to share each node da-
tabase, containing the accumulated traffic knowledge.
Once new information is received about road segment
on the navigational path, one can use our simple but
yet efficient re-routing algorithm.
The dynamic routing application in this project is
the test-bed for our innovative contribution: we pro-
HACIT: A Privacy Preserving and Low Cost Solution for Dynamic Navigation and Forensics in VANET
459
pose a complete and secure decentralized system that
preserves user privacy, since no computation is made
at a central server that could gather private data. In
addition, the system enables forensics logging within
a network made exclusively of mobile device: by use
of data hash stored in a lightweight distributed block-
chain through Smart Contract, we allow each user to
be able to prove its own data integrity. Furthermore,
this blockchain is used to retrieve nodes public keys
and thus, authenticate and decrypt packets.
5.1 Discussion on Mobile Computing
Given the nature of our proposed solution, we foresee
several problems and points that must be addressed:
Given the communication window between no-
des, the data exchanged must be as concise as pos-
sible. To that end, we can prune the data to be
shared to include only latest gathered road infor-
mation.
The computation of the navigation path might be
a challenge for mobile devices. However, nowa-
days smartphones can handle heavy computatio-
nal burden and finding the shortest path is a well
known and optimized algorithm.
We encrypt the data exchange between nodes with
asymmetric encryption. Such encryption is howe-
ver computationally expensive.
The frequency of smart contract execution and
ledger update will directly affect the efficiency of
the proposed solution but also the power and com-
putational requirements.
5.2 Future Work
The solution that we propose is yet to be developed.
The work will question different sensitive subjects
specific to VANET limitations. For instance, we fo-
resee that the size of the database to be shared over a
really small amount of time might be a challenge, or
that the computing power available on mobile device
might not be sufficient to run Dijkstra or asymme-
tric cryptography. The latest project break-in invol-
ves a complete different approach. Indeed, the hyper-
ledger Fabric functionalities allow to store key/value
pairs alongside the ledger of transactions as permissi-
oned blockchain. A user would then directly submit
a new transaction containing a new measured speed
for a certain road id as transaction. Such scheme ena-
ble forensics through transactions crawling and dyn-
amic rerouting through database query. However, the
scalability of the system for mobile application is yet
unknown.
REFERENCES
Beresford, A. R. and Stajano, F. (2004). Mix zones: User
privacy in location-aware services. In Pervasive Com-
puting and Communications Workshops, 2004. Pro-
ceedings of the Second IEEE Annual Conference on,
pages 127–131. IEEE.
Brecht, W. (2012). White-box cryptography: hiding keys in
software. NAGRA Kudelski Group.
Dijkstra, E. W. (1959). A note on two problems in connex-
ion with graphs. Numerische mathematik, 1(1):269–
271.
Directive, E. (1995). 95/46/ec of the european parliament
and of the council of 24 october 1995 on the protection
of individuals with regard to the processing of perso-
nal data and on the free movement of such data. Offi-
cial Journal of the EC, 23(6).
EU (2016). Directive (EU) 2016/680 of the Eu-
ropean Parliament and of the Council of 27
April 2016. http://eur-lex.europa.eu/legal-
content/EN/ALL/?uri=CELEX:32016L0680.
Funai, C., Tapparello, C., and Heinzelman, W. (2015). Sup-
porting multi-hop device-to-device networks through
wifi direct multi-group networking. arXiv preprint
arXiv:1601.00028.
Garip, M. T., Gursoy, M. E., Reiher, P., and Gerla, M.
(2015). Scalable reactive vehicle-to-vehicle conges-
tion avoidance mechanism. In Consumer Communica-
tions and Networking Conference (CCNC), 2015 12th
Annual IEEE, pages 943–948. IEEE.
GoTenna (2016). GoTenna. www.gotenna.com. [Online;
accessed 21-November-2017].
Grignard, A., Taillandier, P., Gaudou, B., Vo, D. A., Huynh,
N. Q., and Drogoul, A. (2013). Gama 1.6: Advan-
cing the art of complex agent-based modeling and si-
mulation. In International Conference on Principles
and Practice of Multi-Agent Systems, pages 117–131.
Springer.
Hadiwardoyo, S. A., Patra, S., Calafate, C. T., Cano, J.-C.,
and Manzoni, P. (2017). An android its driving safety
application based on vehicle-to-vehicle (v2v) commu-
nications. In Computer Communication and Networks
(ICCCN), 2017 26th International Conference on, pa-
ges 1–6. IEEE.
Kaul, A., Obraczka, K., Santos, M., Rothenberg, C., and
Turletti, T. (2017). Dynamically distributed net-
work control for message dissemination in its. In
IEEE/ACM DS-RT 2017-21st International Sympo-
sium on Distributed Simulation and Real Time Appli-
cations.
Leiding, B., Memarmoshrefi, P., and Hogrefe, D. (2016).
Self-managed and blockchain-based vehicular ad-hoc
networks. In Proceedings of the 2016 ACM Interna-
tional Joint Conference on Pervasive and Ubiquitous
Computing: Adjunct, pages 137–140. ACM.
Leontiadis, I., Marfia, G., Mack, D., Pau, G., Mascolo, C.,
and Gerla, M. (2011). On the effectiveness of an op-
portunistic traffic management system for vehicular
networks. IEEE Transactions on Intelligent Transpor-
tation Systems, 12(4):1537–1548.
VEHITS 2018 - 4th International Conference on Vehicle Technology and Intelligent Transport Systems
460
linux Foundation, T. (2016). HyperLedger Fabric.
https://hyperledger-fabric.readthedocs.io/en/release/.
[Online; accessed 21-November-2017].
Melani (2006). MELANI, Reporting and Ana-
lysis Centre for Information Assurance.
https://www.melani.admin.ch/melani/en/home.html.
[Online; accessed 21-November-2017].
Prasad, A., Kunz, A., Velev, G., Samdanis, K., and Song, J.
(2014). Energy-efficient d2d discovery for proximity
services in 3gpp lte-advanced networks: Prose disco-
very mechanisms. IEEE vehicular technology maga-
zine, 9(4):40–50.
Raya, M. and Hubaux, J.-P. (2007). Securing vehicu-
lar ad hoc networks. Journal of computer security,
15(1):39–68.
Serval project (2016). Serval project.
http://www.servalproject.org/. [Online; accessed
21-November-2017].
Sharma, P. K., Moon, S. Y., and Park, J. H. (2017). Block-
vn: A distributed blockchain based vehicular network
architecture in smart city. Journal of information pro-
cessing systems, 13(1):184–195.
Tornell, S. M., Patra, S., Calafate, C. T., Cano, J.-C.,
and Manzoni, P. (2015). Grcbox: extending smartp-
hone connectivity in vehicular networks. Internatio-
nal Journal of Distributed Sensor Networks, 2015.
Veins (2011). VEINs simulator. veins.car2x.org/. [Online;
accessed 21-November-2017].
Wood, G. (2014). Ethereum: A secure decentralised ge-
neralised transaction ledger. Ethereum Project Yellow
Paper, 151.
HACIT: A Privacy Preserving and Low Cost Solution for Dynamic Navigation and Forensics in VANET
461